Automatic power generation control method and equipment based on base station standby battery virtual power plant

Abstract

The invention relates to an automatic power generation control method and equipment based on a base station standby battery virtual power plant. The automatic power generation control method comprises the following steps: S1, acquiring a real-time state vector of each 5G base station; s2, adopting a frequency modulation signal conversion method based on virtual clearing to obtain a unified control signal in a virtual power plant of the standby battery of the base station, and specifically, obtaining an AGC frequency modulation signal, and calculating the total response power of the VPP; constructing virtual bidding vectors of the 5G base stations, and aggregating the virtual bidding vectors into a stepped virtual bidding curve; the intersection point of the virtual bidding curve and the total response power is solved, and a unified control signal in the B3VPP is obtained; and S3, converting a unified control signal in the B3VPP into the actual response power of each 5G base station standby battery through a response algorithm, adjusting the charging and discharging states of each standby battery, and coordinating and controlling automatic power generation. Compared with the prior art, accurate response of the B3VPP to the AGC frequency modulation signal is realized through a distributed control mechanism, and the standby power demand of communication equipment in each base station is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of automatic power generation control, and in particular, relates to an automatic power generation control method and device based on a base station backup battery virtual power plant. Background Art

[0002] In the new power system, as the new energy power generation system will gradually replace the conventional power source, the total inertia of the system will continue to decrease, affecting the frequency regulation capability of the system; at the same time, the volatility of the new energy output increases the demand for fast frequency regulation resources in the power grid. These will bring serious challenges to the traditional automatic generation control (AGC) that mainly relies on conventional units. Battery energy storage can respond to regulation instructions quickly and accurately, and has become a competitive fast frequency regulation resource. However, the deployment of energy storage for frequency regulation requires a lot of investment, and tapping existing energy storage resources is a more economical measure. 5G communication base station backup batteries are an important type of user-side energy storage resources. If these long-term idle energy storage resources can be effectively activated, it can provide an effective solution to the two problems of insufficient fast frequency regulation resources in the power grid and high electricity costs of base stations.

[0003] Base station backup batteries belong to distributed energy storage. The literature Exploring the Cellular Base Station Dispatch Potential Towards Power System Frequency Regulation (YONG Pei, ZHANGNing, LIU Yuxiao, et al., IEEE Transactions on Power Systems, 2022, Vol. 37(1): 820-823) and Robust Control Scheme for Distributed Battery Energy Storage Systems in Load Frequency Control (OSHNOEI A, KHERADMANDI M, MUYEEN S M., IEEE Transactions on Power Systems, 2020, 35(6): 4781-4791) proposed control schemes for distributed energy storage to participate in AGC based on the consistency principle and model predictive control method respectively; the literature Coordinated Control Scheme for Provisionof Frequency Regulation Service by Virtual Power Plants (OSHNOEI A, KHERADMANDIM, BLAABJERG F, et al., Applied Energy, 2022, 325:119734) aggregates distributed energy storage and heat pumps into a virtual power plant (VPP) to participate in AGC, and adjusts power within the VPP based on response speed and available capacity allocation. However, the above method has two shortcomings. First, it cannot analyze the economic feasibility of 5G base stations participating in AGC; second, it does not take into account the special needs of 5G base station batteries that must ensure backup requirements.

[0004] In view of the above-mentioned particularities of 5G base stations, researchers have proposed a variety of coordinated control methods for backup batteries in recent years, and applied them to scenarios such as demand response, distribution network congestion management, and economic dispatch. However, the time scale of the above applications is usually 1 hour, which is not sensitive to the real-time, lightweight, and communication delay of the control. At the same time, applications such as peak shaving and congestion management are not implemented on a regular basis, so little attention is paid to the sustainability of battery SOC, which is crucial for 5G base stations to participate in AGC. Therefore, it is necessary to design an automatic power generation control method for aggregating large-scale 5G base station backup batteries to reduce the dispatching cost of the power grid for distributed energy storage, realize the precise response of the base station backup battery virtual power plant to the automatic power generation control frequency modulation signal, and at the same time, ensure the backup power demand of communication equipment in each base station. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide an automatic power generation control method and device based on a base station backup battery virtual power plant, so as to realize the accurate response of the base station backup battery virtual power plant to the automatic power generation control frequency modulation signal through a distributed control mechanism, and at the same time, fully guarantee the backup power demand of the communication equipment in each base station.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] The present invention provides an automatic power generation control method based on a base station backup battery virtual power plant, which is applicable to an automatic power generation control system in which 5G base station backup batteries participate in coordination. The system includes a power grid dispatching center and a base station backup battery virtual power plant. The base station backup battery virtual power plant includes multiple 5G base stations, each of which includes communication equipment and corresponding backup batteries. The battery capacity of each backup battery is subject to a backup power priority constraint condition to ensure the backup power demand of the corresponding communication equipment.

[0008] The method for automatic power generation control by the above system comprises the following steps:

[0009] S1. Obtain the real-time state vector Ω of each 5G base station i,t , the real-time state vector Ω i,t Including backup battery state of charge indicator S i,t 、Current charge and discharge status of the backup battery Z i,t , Maximum charging power of backup battery and maximum discharge power of backup battery Among them, Z i,t The values ​​are -1, 1 and 0, which represent charging, discharging and idle respectively. The maximum charging power of the backup battery The value is the rated charging power of the backup battery and the maximum discharge power of the backup battery The value is the smaller one between the rated discharge power of the backup battery and the power of the communication equipment;

[0010] S2. Use the frequency modulation signal transformation method based on virtual clearing to obtain the unified control signal inside the base station backup battery virtual power plant The specific process is:

[0011] S201, obtaining the automatic generation control frequency modulation signal R sent by the power grid dispatching center t , calculate the total response power P of the virtual power plant VPP,t ;

[0012] S202: Based on the real-time state vector Ω of each 5G base station i,t , a virtual bidding strategy is used to construct the virtual bidding vector (p i,t ,q i,t ), where p i,t is the virtual bidding price, which is used to indicate the priority of backup battery charging and discharging. i,t is the bid quantity, which is used to indicate the charge and discharge power of the backup battery;

[0013] S203: Based on the delayed aggregation strategy, the virtual bidding vector (p i,t ,q i,t ) are aggregated into a step-type virtual bidding curve;

[0014] S204, obtain virtual bidding curve and total response power P VPP,t The intersection of the base station backup battery virtual power plant obtains the unified control signal inside

[0015] S3, through the response algorithm to the base station backup battery virtual power plant internal unified control signal Transformed into the actual response power P of each 5G base station backup battery i,t , based on which the charge and discharge status of each 5G base station backup battery is adjusted, and automatic power generation is coordinated and controlled.

[0016] Furthermore, the expression of the backup power priority constraint condition is specifically as follows:

[0017]

[0018] Among them, E i,t is the battery capacity of the backup battery in the i-th 5G base station at time t, E max,i is the upper limit of battery power, s is the upper limit coefficient of battery power, E N,i is the rated capacity of the backup battery, E min,i,t is the minimum reserve capacity of the backup battery at time t, is the average communication load of the Γ period predicted a day ago, Γt is the time period to which time t belongs, T b The minimum standby time.

[0019] Furthermore, the backup battery charge state indicator S i,t The specific expression is as follows:

[0020]

[0021] Among them, E i,t is the battery capacity of the backup battery in the i-th 5G base station at time t, E max,i is the maximum value of the battery power, E min,i,t is the minimum backup power of the backup battery at time t.

[0022] Furthermore, in step S201, the total response power P of the virtual power plant VPP,t The calculation formula is as follows:

[0023]

[0024] in, is the base power of the virtual power plant, is the frequency regulation capacity of the virtual power plant, which is calculated by the following formula:

[0025]

[0026] in, and are the rated charge and discharge power of the backup battery respectively, is the average communication load in period Γ predicted on the previous day.

[0027] Furthermore, the base power of the virtual power plant Regularly adjust using the following formula:

[0028]

[0029] Among them, γ base is the base point power adjustment coefficient, For backup battery virtual power plant in Γ t Average backup battery state of charge index value for a period of time, S ideal It is the ideal backup battery state of charge index value of the backup battery virtual power plant.

[0030] Furthermore, in step S202, a price offset is introduced into the virtual bidding strategy, specifically: if the battery is in a non-charging state in the previous cycle, its virtual bidding price is offset by 1 to reduce the charging priority.

[0031] Furthermore, the virtual bidding strategy is as follows:

[0032] When the charge and discharge state of the backup battery in the previous control cycle is Z i,t-1 =-1, if the total response power P VPP,t <0, then p i,t =-S i,t , If the total response power P VPP,t >0, then p i,t =S i,t ,

[0033] When the charge and discharge state of the backup battery in the previous control cycle is Z i,t-1 =0, if the total response power P VPP,t <0, then p i,t =-(S i,t +1), If the total response power P VPP,t >0, then p i,t =S i,t ,

[0034] When the charge and discharge state of the backup battery in the previous control cycle is Z i,t-1 =1, if the total response power P VPP,t <0, then p i,t =-(S i,t +1), If the total response power P VPP,t >0, then p i,t =S i,t +1,

[0035] Furthermore, the delayed aggregation strategy is specifically as follows:

[0036] The virtual bidding vector (p i,t ,q i,t ) is aggregated once per minute and completed in parallel with step S204 and step S3, and both step S204 and step S3 are performed in a cycle of 4 seconds.

[0037] Furthermore, in step S3, the response algorithm is specifically as follows:

[0038] If the base station backup battery virtual power plant has a unified control signal And virtual bidding price The backup battery is charged; if the base station backup battery virtual power plant has a unified control signal inside And virtual bidding price The backup battery discharges; in other cases, charging and discharging are stopped; the power of charging and discharging of the backup battery is taken as the bid quantity q i,t .

[0039] The present invention also provides an electronic device, comprising a memory, a processor, and a program stored in the memory, wherein the processor implements the above method when executing the program.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention proposes an automatic power generation control method based on a base station backup battery virtual power plant, and constructs an automatic power generation control system in which 5G base station backup batteries participate in coordination. The system includes a power grid dispatching center and a base station backup battery virtual power plant. The base station backup battery virtual power plant includes multiple 5G base stations. Each 5G base station includes communication equipment and corresponding backup batteries. The battery capacity of each backup battery is subject to the backup power priority constraint condition, which can guarantee the backup power demand of the corresponding communication equipment. The method for automatic power generation control through the above system includes the following steps: first, obtaining the real-time state vector of each 5G base station; then, using the frequency modulation signal transformation method based on virtual clearing to obtain the unified control signal inside the base station backup battery virtual power plant, the specific process is: first, obtaining the automatic power generation control frequency modulation signal issued by the power grid dispatching center, and calculating the total response power of the virtual power plant; secondly, according to the real-time state vector of each 5G base station, using the virtual bidding strategy to construct the virtual bidding vector of each 5G base station; thirdly, based on the delayed aggregation strategy, the virtual bidding vectors of each 5G base station are aggregated into a step-type virtual bidding curve; then, the intersection of the virtual bidding curve and the total response power is obtained to obtain the unified control signal inside the base station backup battery virtual power plant; after completing the above steps, the unified control signal inside the base station backup battery virtual power plant is transformed into the actual response power of each 5G base station backup battery through the response algorithm, and the charge and discharge state of each 5G base station backup battery is adjusted accordingly; the above method can realize the base station backup battery virtual power plant to accurately respond to the automatic power generation control frequency modulation signal as a whole through a distributed control mechanism, and at the same time, fully guarantee the backup power demand of each base station communication equipment.

[0042] 2. In the frequency modulation signal transformation method based on virtual clearing proposed in the present invention, the base point power of the virtual power plant is adjusted regularly, so that the charge state index values ​​of the backup batteries of each base station can be made consistent and restored to near the ideal value; a price offset is introduced in the virtual bidding strategy. If the battery is in a non-charging state in the previous cycle, its virtual bidding price is offset by 1 to reduce the charging priority. The above setting can minimize the frequency of backup battery state transitions, which is conducive to giving full play to its performance advantage in responding to fast frequency modulation signals; in addition, a delayed aggregation strategy is designed to reduce the impact of uplink communication delays in the virtual power plant, and can significantly reduce the communication and calculation amount of the aggregation link, so that the process of the virtual power plant responding to automatic control signals is closer to that of conventional units.

[0043] 3. On the one hand, the response algorithm of the present invention can ensure that there will be no backup batteries in the charging and discharging states at the same time in the virtual power plant, thereby avoiding unnecessary energy loss; on the other hand, the power of charging and discharging the backup battery is taken as the bidding amount. According to the virtual bidding strategy, the bidding amount depends on the backup battery charge state index. Therefore, the present invention allocates the frequency modulation power according to the battery charge state index, so that each base station can obtain charging and discharging opportunities fairly. In the process of alternating positive and negative changes in the automatic power generation control signal, the backup battery charge state index values ​​of each base station can be made to converge. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The structure diagram of the automatic power generation control system coordinated by the backup battery of the base station;

[0045] Figure 2 This is a schematic diagram of the adjustable capacity of the backup battery.

[0046] Among them, (2a) is the base station communication load prediction curve, (2b) is the schematic diagram of the relationship between the adjustable capacity and the backup capacity of the backup battery, is the average communication load of the Γ period predicted a day ago, S is the charge state indicator of the backup battery;

[0047] Figure 3 It is a step-type virtual bidding curve;

[0048] Among them, (3a) corresponds to VPP charging, (3b) corresponds to VPP discharging, P VPP,t is the total response power, is the base power of the virtual power plant, is the frequency regulation capacity of the virtual power plant, R t It is the AGC frequency modulation signal;

[0049] Figure 4 This is a schematic diagram of the B3VPP delayed aggregation strategy.

[0050] Among them, (4a) corresponds to the clearing-execution phase (with a period of 4 seconds), and (4b) corresponds to the aggregation phase (with a period of 1 minute);

[0051] Figure 5 are the typical load curves of four base stations.

[0052] Among them, (5a), (5b), (5c), and (5d) are typical load curves of a base station respectively;

[0053] Figure 6 The calculation results of the frequency regulation capacity of B3VPP in each period;

[0054] Figure 7 is a schematic diagram of communication delay distribution.

[0055] Among them, (7a) is the downlink communication delay between the power grid and the aggregator, (7b) is the downlink communication delay between the aggregator and the base station, and (7c) is the uplink communication delay between the base station and the aggregator;

[0056] Figure 8 is the daily charge and discharge state switching times distribution of the base station backup battery,

[0057] Among them, (8a) is the number of charge and discharge state switching times of scenario 1, and (8b) is the number of charge and discharge state switching times of scenario 2;

[0058] Fig. 9 The percentage of base stations whose battery status changes in each AGC cycle,

[0059] Among them, (9a) corresponds to scenario 1, and (9b) corresponds to scenario 2;

[0060] Fig.10 It is a schematic diagram of tracking delay of AGC signal;

[0061] Fig.11 This is a schematic diagram of the effect of VC-RST.

[0062] Among them, (11a) is the response RegD signal, and (11b) is the response RegA signal;

[0063] Fig.12 Schematic diagram of base point power adjustment effect (response RegD),

[0064] Among them, (12a) is the change of VPP average SOC, (12b) is the change of VPPS value;

[0065] Fig.13 This is a schematic diagram of the SOC change when considering a base station power outage.

[0066] Among them, curves of different colors correspond to different backup batteries in the power outage base station. DETAILED DESCRIPTION

[0067] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0068] Example 1

[0069] At present, most of my country's 5G macro base stations are built and operated by China Tower Corporation. The communication equipment in the base station (including baseband units, active antenna units, etc.) belongs to the communication operators, while the rest of the equipment, including backup batteries, belongs to the tower company. Therefore, the tower company has the authority to monitor the power of the base station and regulate the backup batteries, and is thus qualified to build a virtual power plant (VPP of Base station Backup Batteries, B3VPP) as an operator and participate in the automatic generation control (AGC) auxiliary service.

[0070] This embodiment provides an automatic power generation control method based on a base station backup battery virtual power plant, which is applicable to an automatic power generation control system in which a base station backup battery participates in coordination, such as Figure 1 As shown, the system includes a power grid dispatching center and B3VPP, and B3VPP includes multiple 5G base stations, each of which includes communication equipment and corresponding backup batteries. This embodiment adopts the idea of ​​"time division multiplexing" to improve the utilization rate of the base station backup battery, that is, when the base station is normally powered by the mains through the switching power supply, the backup battery participates in the AGC auxiliary service; and when the base station loses the mains, the battery immediately switches to the backup power state to ensure the continuous and uninterrupted operation of the communication equipment.

[0071] The power grid dispatcher generally sends out AGC frequency modulation signals with a period of 4 seconds, denoted as R t ∈[-1,1]. B3VPP needs to coordinate a large number of base stations to accurately respond to the signal, which can be described by a mathematical formula:

[0072]

[0073] Among them, f VPP () represents the decomposition algorithm of B3VPP, which aims to t Forming a unified control signal inside VPP Function g i () represents the response algorithm of each base station, which is responsible for Transformed into the actual response power P of the base station backup battery i,t ; The subscript i represents different base stations. B3VPP relies on broadcasting a unified signal to each base station Distributed control is implemented without sending individual adjustment commands (which corresponds to centralized control), thus adapting to the scale characteristics of 5G base stations.

[0074] The above decomposition and response algorithm must meet the following constraints:

[0075] 1) The VPP as a whole should accurately respond to the FM signal, that is, it should meet the following requirements:

[0076]

[0077] In the above formula, M is the total number of base stations in the VPP, P VPP,t is the total response power (aggregate power), is the base power of the virtual power plant, is the frequency regulation capacity of the virtual power plant, Then is the frequency modulation power that the VPP needs to respond to, Γ t is the time period to which time t belongs, and the power is positive when the battery is discharging.

[0078] 2) Each base station should maintain the operational feasible domain of its backup battery. The charging and discharging power of the backup battery should meet the following requirements:

[0079]

[0080] In the above formula, and are the battery charging and discharging power, and are the rated charge and discharge power, P com,i,t is the base station communication equipment power, is the maximum charging power of the backup battery, is the maximum discharge power of the backup battery. Formula (3c) is used to ensure that the switching power supply of the base station will not reverse power, and formula (3d) is the mutually exclusive constraint of the battery charging and discharging state.

[0081] The relationship between the power and capacity of the backup battery is:

[0082]

[0083] In the above formula, E i,t is the battery capacity of the backup battery in the i-th 5G base station at time t, Δt is the time interval, η cha and η dch are the battery charging and discharging efficiency respectively.

[0084] Unlike conventional generators, backup batteries are limited energy resources. When continuously participating in frequency regulation, their remaining power SOC should always meet the following requirements:

[0085]

[0086] In the above formula, E max,i is the upper limit of battery power, s is the upper limit coefficient of battery power, E N,i is the rated capacity of the backup battery, is the average communication load of the Γ period predicted a day ago, such as Figure 2 As shown in (2a), T b is the minimum standby time, E min,i,t is the minimum reserve power of the backup battery at time t, such as Figure 2 as shown in (2b).

[0087] Since the grid frequency modulation signal R t ∈[-1,1], so the power grid usually hopes that the AGC resources provide positive and negative symmetrical regulation capabilities. The present invention estimates the frequency regulation capacity of B3VPP in each time period according to the following formula:

[0088]

[0089] The present invention achieves the above requirements and performs AGC control by the following steps:

[0090] S1. Obtain the real-time state vector Ω of each 5G base station i,t , the real-time state vector Ω i,t It consists of the following four tuples:

[0091]

[0092] Among them, S i,t is the charge state indicator of the backup battery relative to the adjustable capacity and is defined as follows:

[0093]

[0094] Z i,t The current charge and discharge status of the backup battery. The values ​​are -1, 1, and 0, representing charging, discharging, and idle, respectively.

[0095] To ensure that the variables in constraints (2)-(5) are consistent with the actual values, the real-time state vector of the backup battery is periodically updated through step S1.

[0096] S2, using the virtual clearing-based frequency modulation signal transformation method (VC-RST) to obtain the unified control signal inside the base station backup battery virtual power plant The specific process is:

[0097] S201, B3VPP obtains the AGC signal R sent by the power grid dispatching center in real time t , calculate the total response power P of the virtual power plant according to formula (2) VPP,t .

[0098] Due to the non-zero mean value of the AGC signal and the battery's own charge and discharge losses, it is necessary to replenish the backup battery according to a certain strategy to restore its S value to near the ideal value. Therefore, this embodiment proposes a VPP base point power adjustment strategy to regularly adjust the power consumption of the backup battery. Make adjustments.

[0099] First, the S value of B3VPP at time t is defined as:

[0100]

[0101] Then, B3VPP is in Γ t The average S value of a period is defined as:

[0102]

[0103] In the above formula, |Γ t | is the number of calculations of formula (9) in a period.

[0104] Finally, the base point power of VPP is calculated as follows:

[0105]

[0106] In the above formula, γ base is the base point power adjustment coefficient, S ideal is the ideal standby battery state of charge index value of the standby battery virtual power plant, which is 0.5 in this embodiment.

[0107] In this embodiment, the base point power is adjusted once every hour according to formula (11), and the base point power is added to the real-time frequency modulation power according to formula (2). The total power obtained is finally redistributed to each base station by the VC-RST method.

[0108] S202: Based on the real-time state vector Ω of each 5G base station i,t , a virtual bidding strategy is used to construct the virtual bidding vector (p i,t ,q i,t ), where p i,t is the virtual bidding price, which is used to indicate the priority of backup battery charging and discharging. i,t It is the bid quantity, which is used to indicate the charge and discharge power of the backup battery.

[0109] Energy storage power control allows a dynamic response time of 2s and an adjustment time of 3s, which is close to the AGC signal cycle. Therefore, minimizing the frequency of battery state transitions is conducive to giving full play to its performance advantage in responding to fast frequency modulation signals. To achieve the above purpose, the present invention introduces price offset in the virtual bidding strategy. If the battery is in a non-charging state in the previous control cycle, its virtual bidding price is offset by 1 to reduce the charging priority. The virtual bidding strategy is specifically shown in Table 1:

[0110] Table 1 Virtual bidding strategy

[0111]

[0112] S203: The virtual bidding vector (p i,t ,q i,v ) are aggregated into a step-type virtual bidding curve.

[0113] VPP is based on the total response power P VPP,t The symbol of each base station is p i,t The order from high to low is aggregated into Figure 3 The step-shaped virtual bidding curve is shown in Figure 1, where (3a) corresponds to VPP charging and (3b) corresponds to VPP discharging. Each line segment in the bidding curve corresponds to a virtual bidding vector (p i,t ,q i,t ), whose color indicates the charging and discharging status of the base station battery in the previous control cycle.

[0114] S204, obtain virtual bidding curve and total response power P VPP,t The intersection of the base station backup battery virtual power plant is marked as the unified control signal

[0115] S3, through the response algorithm to the base station backup battery virtual power plant internal unified control signal Transformed into the actual response power P of each 5G base station backup battery i,t , and accordingly adjust the charge and discharge status of each 5G base station backup battery, and coordinate and control automatic power generation. The response algorithm is as follows:

[0116] If the base station backup battery virtual power plant has a unified control signal And virtual bidding price The backup battery is charged; if the base station backup battery virtual power plant has a unified control signal inside And virtual bidding price The backup battery discharges; the power of charging and discharging the backup battery is taken as the bid quantity q i,tIn other cases, charging and discharging are stopped. According to the above response algorithm, there will not be batteries in the charging and discharging states at the same time in the VPP, thus avoiding unnecessary energy loss. On the other hand, VC-RST allocates frequency modulation power according to the S value of the battery, so that each base station can get charging and discharging opportunities fairly. In the process of alternating positive and negative changes of the AGC signal, such a coordination strategy will lead to the convergence of the S values ​​of each base station.

[0117] It can be seen from the above that the VC-RST method includes three steps, namely aggregation (including step S1, step S201, step S202 and step S203), virtual clearing (step S204) and execution (step S3). Compared with conventional units, B3VPP is more susceptible to communication delays when participating in AGC. On the one hand, this is because B3VPP needs to aggregate large-scale base stations, and on the other hand, it is because power-dedicated networks are usually not available in VPP. Taking into account the impact of communication delays, the present invention considers two aggregation strategies: 1) Synchronous aggregation: VPP performs aggregation in each AGC cycle, and completes it in series with the virtual clearing and execution steps; 2) Lazy aggregation: such as Figure 4 As shown in (4b), VPP aggregates only once per minute and is completed in parallel with the virtual clearing and execution phases, as shown in Figure 4 As shown in (4a), the virtual clearing and execution phases still have a cycle of 4 seconds.

[0118] Depend on Figure 4 It can be seen that the delayed aggregation strategy can reduce the impact of uplink communication delay in VPP, and can significantly reduce the communication and calculation amount of the aggregation link, making the process of VPP responding to AGC signals closer to conventional units. The disadvantage of delayed aggregation is that it will cause each base station to fail to update the S value of the battery in time. However, even if the battery is charged and discharged at a 1C rate, its SOC change within 1 minute is less than 2%, which has little impact on control accuracy.

[0119] The following example takes the AGC signal as negative to illustrate how the VC-RST algorithm satisfies equations (1)-(5). At this time, B3VPP needs to be charged to achieve downward regulation, corresponding to the second column of Table 1 and Figure 3 Step S2 implements the function f in equation (1). VPP (), R t Transformed into a unified control signal inside VPP Depend on Figure 3 As can be seen from (3a), the battery with a lower S value (relative SOC after deducting the reserve capacity) will be charged first, which will make all batteries more likely to remain in their original state (continue to charge or idle), maintaining their energy constraint, that is, formula (5); It can be regarded as the equilibrium price, so its coordination result satisfies formula (2). Step S3 implements the function g in formula (1)i (),Will Converted into the actual control power P of the battery i,t , the backup battery determines the bidding amount according to formula (3) to ensure its power constraint.

[0120] To prove the effectiveness of the above method, this embodiment establishes a benefit model of B3VPP participating in AGC and further verifies it through simulation examples.

[0121] The benefit model of B3VPP participating in AGC includes the following two parts:

[0122] Ⅰ.B3VPP’s service costs for participating in AGC

[0123] The additional cost incurred by the base station due to participating in AGC is called service cost, which is composed of the power cost C energy,t , Battery life loss cost C life-cost,t and additional investment cost C ex-cap Three components.

[0124] (1) Electricity cost C energy

[0125]

[0126] In the above formula, p e,t The total electricity cost in a day is mainly affected by the change of electricity price, while the charging and discharging loss will also potentially increase the electricity cost.

[0127] (2) Battery life loss cost C life-cost,t

[0128] The loss of battery life is mainly related to the number of charge and discharge times and the depth of discharge (DOD). This embodiment calculates the equivalent full cycle number of the battery's current action according to the following formula:

[0129]

[0130] In the above formula, N 100％ is the number of cycles of the battery under 100% DOD, d t is the DOD value of this charge and discharge, and a and b are fitting coefficients.

[0131] Then use the annuity method to calculate the battery life loss cost after this action:

[0132]

[0133] In the above formula, C Ess is the purchase cost of the battery system, T Ess-life is the expected service life of the battery, and r is the discount rate.

[0134] (3) Additional investment cost C ex-cap

[0135] Participating in AGC requires additional investment in a real-time power meter system, with the cost calculated at 200 yuan per base station. The additional investment cost per base station per day is calculated as follows:

[0136]

[0137] In the above formula, T met-life It is the expected operating life of the power meter system.

[0138] II.B3VPP’s Benefits of Participating in AGC

[0139] (1) Frequency modulation benefits of the domestic “two rules”

[0140] As the AGC ancillary service market matures, the compensation method for frequency regulation units has evolved from a single compensation based on frequency regulation power to a market transaction compensation based on frequency regulation performance. This embodiment takes the North China Power Grid (see the North China Energy Regulatory Bureau, Notice on Soliciting Opinions on the "Implementation Rules for the Management of Power Grid Connection Operation in the North China Region" and the "Implementation Rules for the Management of Power Auxiliary Services in the North China Region" (Draft for Comments)) as an example, and divides the unit frequency regulation compensation income into frequency regulation capacity compensation BNC CAP and frequency regulation power compensation:

[0141]

[0142] In the above formula, is the frequency regulation capacity compensation price, which is 10 yuan·(MW·h) in this embodiment. -1 , D is the actual daily adjustment depth, K p It is a comprehensive performance indicator, which is determined by the response time K1, the adjustment rate K2 and the adjustment accuracy K3. AGC is the frequency modulation power compensation price, which is 10 yuan·(MW·h) in this embodiment. -1 .

[0143] (2) Frequency regulation revenue in the U.S. PJM market

[0144] In order to enable conventional units and energy storage to jointly participate in AGC services, the US PJM decomposes the frequency regulation signal into a conventional signal RegA and a high-frequency signal RegD. The former is responded by conventional units to reduce unit regulation losses; the latter is responded by fast resources such as energy storage to improve frequency regulation performance. At the same time, since the high-frequency RegD component has an average value of 0 within a certain period of time, the capacity requirements of energy storage can be significantly reduced.

[0145] Frequency regulation revenue in PJM frequency regulation market is divided into capacity revenue and adjust mileage earnings

[0146]

[0147] In the above formula, p cap is the capacity compensation price ($ / (MW·h)), p mileage is the mileage compensation price, M ratio is the mileage ratio, S PJM is the frequency modulation response performance index, according to the accuracy coefficient S p , correlation coefficient S corr and the delay coefficient S delay calculate.

[0148] The following is a specific example analysis:

[0149] Ⅰ. Example Setup

[0150] The AGC signal (including RegD and RegA) of the present invention adopts the actual data of the PJM market in the United States in October 2017, and the electricity price adopts the power purchase price of the power grid agent in Shanghai in May 2024. The present invention aggregates 400 5G base stations to form a VPP. The four typical load curves of the base stations are as follows: Figure 5 (5a), (5b), (5c), and (5d) are shown in Table 2. The frequency regulation capacity of B3VPP in each period calculated by formula (6) is as follows: Figure 6 According to the field test, the present invention obtains the communication delay distribution, as shown in Figure 7 As shown, (7a) is the downlink communication delay between the grid and the aggregator, (7b) is the downlink communication delay between the aggregator and the base station, and (7c) is the uplink communication delay between the base station and the aggregator. In addition to the communication delay, the total AGC delay also includes the battery body response delay. The regulation characteristics of the backup batteries of each base station are randomly generated according to Table 3.

[0151] Table 2 Main parameter settings of base station

[0152]

[0153] Table 3 Regulation characteristics of battery energy storage

[0154]

[0155]

[0156] Ⅱ. AGC real-time response performance

[0157] In order to examine the real-time response capability of AGC of B3VPP, this embodiment uses the fast frequency modulation signal RegD and the response performance assessment method of PJM to design two comparison scenarios. Scenario 1: No price offset is introduced in Table 1, and a synchronous aggregation strategy is adopted; Scenario 2: Price offset is introduced, and a delayed aggregation strategy is adopted. The comparison results are shown in Table 4. As can be seen from Table 4, since scenario 2 reduces the number of battery state switching and reduces communication delay, its accuracy coefficient S p and the delay coefficient S delay Improved accordingly.

[0158] Table 4 AGC response performance indicators

[0159]

[0160] The following are analyzed separately:

[0161] A. Number of battery status switching times.

[0162] Depend on Figure 8 From the comparison between (8b) and (8a), it can be seen that after the price offset is introduced in scenario 2, the average daily charge and discharge state switching of the battery is significantly reduced from 7148 times to 498 times; Fig. 9 The comparison between (9b) and (9a) shows that in each AGC cycle, the proportion of base stations with battery charging and battery replacement status changes is significantly reduced. This improves the accuracy coefficient S of scenario 2. p .

[0163] B. Communication delay.

[0164] In each AGC cycle, the average delay from the grid dispatching command to the base station battery response signal is 6.9s and 4.4s for scenario 1 and scenario 2 respectively. Fig.10 It directly reflects the tracking of AGC instructions by B3VPP. As shown in Table 4, the delayed aggregation strategy improves S delay Performance indicators.

[0165] The subsequent simulations in this embodiment are all based on scenario 2, that is, introducing price deviation and adopting a delayed aggregation strategy.

[0166] III. Backup battery charge state maintenance effect

[0167] Next, it is verified whether B3VPP can maintain the backup battery SOC within the limit and ensure the backup power demand in the continuous response to the AGC signal. The above capabilities are jointly guaranteed by the VC-RST method of the present invention and the VPP base point power adjustment strategy.

[0168] A. VC-RST effect.

[0169] VC-RST is used to reasonably distribute the frequency modulation power in VPP. To verify its effect, VPP is made to respond to RegD and RegA signals respectively. Fig.11 visible:

[0170] 1) VC-RST converts the AGC signal into a unified control signal within the VPP Since the virtual bid price is offset by 1, It may be greater than 1. From the actual results, All less than 2. Figure 3 This indicates that the control signal does not cause the S value of the backup battery in VPP to exceed the limit, which is also from Fig.11 The distribution of S values ​​has been verified.

[0171] 2) When responding to the RegD signal, Basically around 1.5, this is because the RegD signal changes frequently and its mean is close to 0, so the S value of each backup battery is maintained around 0.5. The RegA signal does not have positive and negative symmetry, so The S value of the backup battery fluctuates in a larger range. However, judging from the distribution of the backup battery SOC and the backup capacity, the VC-RST method guarantees the backup power demand of all base stations regardless of the response to RegD or RegA.

[0172] 3) The VC-RST method allocates power according to the S value of the backup battery, so that each base station can get a fair chance to charge and discharge. During the alternating positive and negative changes of the AGC signal, the S value of the backup battery of each base station remains approximately the same.

[0173] B. VPP base point power adjustment effect.

[0174] Taking the response to the RegD signal as an example, the results with and without base point power adjustment are compared.

[0175] Depend on Fig.12 It can be seen from (12a) and (12b) that if the base power adjustment strategy is not adopted, even if the mean value of the RegD signal is approximately 0, the average SOC and S value of the backup battery in the VPP will gradually decrease over time due to the charge and discharge loss. However, when the base power adjustment strategy is adopted, the average SOC and S value of the VPP are well maintained.

[0176] C. The base station is out of power.

[0177] According to the statistics of power outages of base stations in a certain urban area, 88.4% of the base stations have an average power outage frequency of no more than 1 time per month, and 90% of the base stations have an average power outage duration of less than 3 hours per month. The power outage probability and duration of the base station are set with reference to the above data, and the response RegD is re-simulated as an example. The results show that the response performance index SPJM is reduced to 0.963 (reduced by 2%). Fig.13 It can be seen that the backup battery of the power-off base station is switched to the backup power state, and its SOC continues to decrease. However, when the power supply is restored, its SOC can be quickly restored and become consistent with other base stations.

[0178] IV. Economic Benefits

[0179] The following is a comparison of the benefits of B3VPP responding to RegD and RegA signals. The former is calculated according to PJM's evaluation indicators, and the latter is calculated according to my country's "two detailed rules". In addition, the benefits of B3VPP participating in peak-valley electricity price arbitrage and quasi-linear demand response are compared. The results are uniformly converted to $ as shown in Table 5. The reason for selecting these two types of applications is that they can be implemented on a regular basis and can be directly compared. However, the peak shaving and valley filling response is implemented only a limited number of times each year, so this embodiment does not make a comparison.

[0180] Table 5 Benefits of B3VPP when participating in different applications ($)

[0181]

[0182] According to Table 5, the following conclusions can be drawn:

[0183] 1) Regarding service costs: B3VPP will pay a higher service cost for participating in AGC, which is mainly composed of the electricity cost caused by changes in electricity prices and the loss of battery cycle life. Peak-valley arbitrage and quasi-linear response can take advantage of electricity price arbitrage, so their electricity cost is negative. In addition, these two types of non-real-time applications do not require the installation of a real-time power meter system, so their additional investment cost is 0.

[0184] 2) Regarding economic compensation: PJM's compensation for energy storage-type fast frequency regulation resources is higher, $300.7 more than the "two detailed rules". As can be seen from Table 5, the main reason is that the frequency regulation capacity compensation in the "two detailed rules" is much lower than the performance compensation, while the PJM market pays more attention to the capacity value of frequency regulation resources. In addition, as can be seen from formula (17), PJM considers the response performance SPJM when calculating capacity compensation and performance compensation. Since the energy storage response RegD has excellent performance advantages, it can obtain higher compensation.

[0185] 3) Regarding VPP net income: Under the current peak-valley price difference and battery cycle number level, peak-valley arbitrage using backup batteries is in a state of micro-profit. In contrast, B3VPP can significantly increase income by participating in quasi-linear response or responding to RegD signals. For example, the latter can increase VPP net income to 7 times that of the peak-valley arbitrage scenario.

[0186] In summary, the backup batteries of 5G base stations have small capacity and dispersed locations. B3VPP can aggregate these resources so that they have external characteristics similar to conventional frequency modulation units, and declare frequency modulation capacity as a whole and track AGC signals. This significantly reduces the grid's dispatching cost for distributed energy storage, and also makes distributed energy storage eligible to participate in AGC auxiliary services. The present invention adopts a frequency modulation signal transformation (VC-RST) method based on virtual clearing and a distributed control mechanism, combined with a delayed aggregation strategy and a base point power adjustment strategy, which can achieve real-time decomposition and accurate response of frequency modulation power, ensure the backup power requirements of each base station, and meet the application requirements of large-scale response of 5G base stations. Example analysis shows that the above method can enable B3VPP to accurately respond to AGC signals and obtain considerable frequency modulation benefits while ensuring the backup power of each base station.

[0187] Example 2

[0188] This embodiment provides an electronic device, including a memory and a processor, wherein the processor is used to execute a program stored in the memory, wherein the program includes a number of instructions and can execute all or part of the steps of the method described in Example 1. The memory includes a computer-readable storage medium, which can be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0189] The above description of the embodiments is to facilitate the understanding and use of the invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments, and improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the present invention should be within the scope of protection of the present invention.

Claims

1. An automatic power generation control method based on a base station backup battery virtual power plant, characterized in that: An automatic power generation control system applicable to 5G base station backup batteries participating in coordination, the system includes a power grid dispatching center and a base station backup battery virtual power plant, the base station backup battery virtual power plant includes multiple 5G base stations, each 5G base station includes communication equipment and corresponding backup batteries, the battery capacity of each backup battery is subject to the backup power priority constraint condition, and the backup power demand of the corresponding communication equipment is guaranteed; The method for automatic power generation control by the above system comprises the following steps: S1. Obtain the real-time state vector Ω of each 5G base station i,t , the real-time state vector Ω i,t Including backup battery state of charge indicator S i,t 、Current charge and discharge status of the backup battery Z i,t , Maximum charging power of backup battery and maximum discharge power of backup battery Among them, Z i,t The values ​​are -1, 1 and 0, which represent charging, discharging and idle respectively. The maximum charging power of the backup battery The value is the rated charging power of the backup battery and the maximum discharge power of the backup battery The value is the smaller one between the rated discharge power of the backup battery and the power of the communication equipment; S2. Use the frequency modulation signal transformation method based on virtual clearing to obtain the unified control signal inside the base station backup battery virtual power plant The specific process is: S201, obtaining the automatic generation control frequency modulation signal R sent by the power grid dispatching center t , calculate the total response power P of the virtual power plant VPP,t ; S202: Based on the real-time state vector Ω of each 5G base station i,t , a virtual bidding strategy is used to construct the virtual bidding vector (p i,t ,q i,t ), where p i,t is the virtual bidding price, which is used to indicate the priority of backup battery charging and discharging. i,t is the bid quantity, which is used to indicate the charge and discharge power of the backup battery; S203: Based on the delayed aggregation strategy, the virtual bidding vector (p i,t ,q i,t ) are aggregated into a step-type virtual bidding curve; S204, obtain virtual bidding curve and total response power P VPP,t The intersection of the base station backup battery virtual power plant obtains the unified control signal inside S3, through the response algorithm to the base station backup battery virtual power plant internal unified control signal Transformed into the actual response power P of each 5G base station backup battery i,t , based on which the charge and discharge status of each 5G base station backup battery is adjusted, and automatic power generation is coordinated and controlled.

2. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 1 is characterized in that: The expression of the backup power priority constraint condition is specifically as follows: Among them, E i,t is the battery capacity of the backup battery in the i-th 5G base station at time t, E max,i is the upper limit of battery power, s is the upper limit coefficient of battery power, E N,i is the rated capacity of the backup battery, E min,i,t is the minimum reserve capacity of the backup battery at time t, is the average communication load of the Γ period predicted a day ago, Γ t is the time period to which time t belongs, T b The minimum standby time.

3. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 1 is characterized in that: The backup battery charge state indicator S i,t The specific expression is as follows: Among them, E i,t is the battery capacity of the backup battery in the i-th 5G base station at time t, E max,i is the maximum value of the battery power, E min,i,t is the minimum backup power of the backup battery at time t.

4. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 1 is characterized in that: In step S201, the total response power P of the virtual power plant VPP，t The calculation formula is as follows: in, is the base power of the virtual power plant, is the frequency regulation capacity of the virtual power plant, which is calculated by the following formula: in, and are the rated charge and discharge power of the backup battery respectively, is the average communication load in period Γ predicted on the previous day.

5. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 4 is characterized in that: Base power of virtual power plants Regularly adjust using the following formula: Among them, γ base is the base point power adjustment coefficient, For backup battery virtual power plant in Γ t Average backup battery state of charge index value for a period of time, S ideal It is the ideal backup battery state of charge index value of the backup battery virtual power plant.

6. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 1 is characterized in that: In step S202, a price offset is introduced into the virtual bidding strategy, specifically: if the battery is in a non-charging state in the previous cycle, its virtual bidding price is offset by 1 to reduce the charging priority.

7. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 6 is characterized in that: The virtual bidding strategy is as follows: When the charge and discharge state of the backup battery in the previous control cycle is Z i,t-1 =-1, if the total response power P VPP,t <0, then p i,t =-S i,t , If the total response power P VPP,t >0, then p i,t =S i,t , When the charge and discharge state of the backup battery in the previous control cycle is Z i,t-1 =0, if the total response power P VPP,t <0, then p i,t =-(S i,t +1), If the total response power P VPP,t >0, then p i,t =S i,t , When the charge and discharge state of the backup battery in the previous control cycle is Z i,t-1 =1, if the total response power P VPP,t <0, then p i,t =-(S i,t +1), If the total response power P VPP,t >0, then p i,t =S i,t +1, 8. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 1 is characterized in that: The delayed aggregation strategy is specifically: The virtual bidding vector (p i,t ,q i,t ) is aggregated once per minute and completed in parallel with step S204 and step S3, and both step S204 and step S3 are performed in a cycle of 4 seconds.

9. The automatic power generation control method based on the base station backup battery virtual power plant according to claim 1 is characterized in that: In step S3, the response algorithm is as follows: If the base station backup battery virtual power plant has a unified control signal And virtual bidding price The backup battery is charged; if the base station backup battery virtual power plant has a unified control signal inside And virtual bidding price The backup battery discharges; in other cases, charging and discharging are stopped; the power of charging and discharging of the backup battery is taken as the bid quantity q i,t .

10. An electronic device comprising a memory, a processor, and a program stored in the memory, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

Images